1. Field of the Invention
The present invention relates to an image processing apparatus, more particularly, an image processing apparatus that displays a locus involved in an image provided by being imaged in time series, an image processing method thereof, and a program adapted for a computer to perform the method.
2. Description of Related Art
Simple synthesis of more than one image (which is hereinafter referred to as a frame image) by overlapping, for instance, is suggested as one method for generating an image indicative of a moving object locus from more than one frame image provided by being imaged in time series. According to the simple synthesis of this type, a rate of single frame contribution to a synthesis result decreases as the number of frames is increased, causing a problem that the moving object locus displayed becomes light-colored. When assuming that there are five frame images as shown in FIG. 10, for instance, the synthesis simply made on these frame images causes the rate of each pixel contribution to the synthesis result to decrease down to 20%. And, in a pixel position at which a moving object traverses, the synthesis is provided in a moving object-to-background ratio of 20:80, leading to an occurrence of light-colored display of the moving object locus as shown in FIG. 11C.
On the other hand, one technique provided by making improvements on the simple synthesis is suggested to synthesize a single still image by, after extracting more than one frame image from video data, giving weighting suited for a luminance value (See Japanese Patent Application Publication Hei. 10-290450 (FIG. 1), for instance). The technique according to this related art provides the synthesis on same positional pixels involved in each frame image by heavily weighting the pixels that are away from a mean luminance value by a distance or more of a dispersion of the luminance values, while lightly weighting the pixels that are away from the mean luminance value by the distance or less of the above dispersion, followed by output of the synthesis result as the same positional pixels of an output image. This technique ensures that in a synthesized image, the moving object is more heavily weighted than a background, causing the moving object locus to be displayed more clearly, as compared with the simple synthesis.
In addition, one different technique is suggested to provide the synthesis of the moving object locus by extracting and tracing the moving object (See Japanese Patent Application Publication 2005-123824 (Refer to FIG. 1), for instance). The technique according to this related art ensures that an enhanced accuracy of extraction and tracing of the moving object is obtained by, when the tracing of the moving object results in a failure, interpolating the moving object based on a moving object characteristic amount and/or position provided at the time of successful tracing.